Core mission: Commitment to openness

The Open Semantic Interchange (OSI) initiative was launched as a collaborative, open source effort to solve one of the most persistent and costly challenges in the data world: inconsistent metric definitions. The initiative is guided by key principles, including: 

• Standardization: Creating a uniform language for definitions
• Interoperability: Facilitating seamless data exchange
• Extensibility: Allowing the model to adapt as data needs evolve 

By championing this common, vendor-agnostic standard, the OSI will help to accelerate AI and BI adoption and streamline data operations. We’re thrilled that this vital mission continues to gain significant momentum, drawing in new participants from across the industry as we work to build this essential framework.

First working group session

On October 17, OSI held its first official working group session. Members joined in person at the Snowflake offices in Menlo Park, California, Bellevue, Washington, and remotely. The energy and unified commitment to building a shared standard were remarkable, marking a pivotal moment for the initiative.

We were excited to have new working group members from AWS, Collibra, DataHub, Domo, Firebolt, Google, Informatica, Instacart, JPMC, Preset, Starburst Data and Strategy. Their immediate involvement underscores the industry-wide recognition of OSI’s critical role in the future of data and AI. These leaders join our existing powerhouse of ecosystem partners including Alation, Atlan, BlackRock, Blue Yonder, Cube, dbt Labs, Elementum AI, Hex, Honeydew, Mistral AI, Omni, RelationalAI, Salesforce, Select Star, Sigma and ThoughtSpot. We had active discussions about a number of crucial areas, including defining the core technical specification, establishing the governance model necessary for the open standard and setting clear milestones for accelerating interoperability across the ecosystem. 

OSI update: Progress on standardization and MetricFlow

Additionally, dbt Labs recently announced at Coalesce that it will open source MetricFlow and contribute it under the Apache 2.0 license. This is an important contribution to OSI and the open source community. The OSI working group will utilize this newly open sourced project as an initial reference implementation. Leveraging an existing open source tool allows the working group to accelerate the process of defining a comprehensive v1.0 standard, providing a solid, open foundation for unified semantic knowledge.

The path to a comprehensive open standard

The working group is committed to maintaining that this specification is truly vendor neutral and comprehensive, to support the most complex enterprise use cases. This evolution includes:

• Enhanced specification: We will be rapidly incorporating advanced semantic features necessary for complex, enterprise-grade AI applications and achieving full, seamless semantic interoperability across platforms.
• Neutrality and extensibility: We are making sure the OSI specification remains vendor neutral, enabling any organization to implement the standard using their preferred semantic engine without fear of vendor lock-in.
• Shared leadership: As the specification matures, OSI founding members are committed to shared leadership and transparent governance, working together to keep the project neutral and open, to benefit the entire data and AI ecosystem for years to come.

Stay tuned for further updates as the OSI takes shape.